Changeset - 3a12df6cbf30
[Not reviewed]
default
0 2 0
Mads Kiilerich (mads) - 6 years ago 2020-06-11 21:15:07
mads@kiilerich.com
Grafted from: d479422fca87
lib: use sha1 instead of md5 in a couple of places

md5 is dead and should be avoided. In the places changed here, we want to keep
using hashes without trivial collisions, but do not expect strong crypto
security. sha1 seems like a trivial step up from md5 and without obvious
alternatives. It is more expensive than md5, but we can live with that in these
places.

The remaining few uses of md5() cannot be changed without breaking backwards
compatibility or external API.
2 files changed with 4 insertions and 4 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/celerylib/__init__.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# This program is free software: you can redistribute it and/or modify
 
# it under the terms of the GNU General Public License as published by
 
# the Free Software Foundation, either version 3 of the License, or
 
# (at your option) any later version.
 
#
 
# This program is distributed in the hope that it will be useful,
 
# but WITHOUT ANY WARRANTY; without even the implied warranty of
 
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 
# GNU General Public License for more details.
 
#
 
# You should have received a copy of the GNU General Public License
 
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
"""
 
kallithea.lib.celerylib
 
~~~~~~~~~~~~~~~~~~~~~~~
 

	
 
celery libs for Kallithea
 

	
 
This file was forked by the Kallithea project in July 2014.
 
Original author and date, and relevant copyright and licensing information is below:
 
:created_on: Nov 27, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 

	
 
import logging
 
import os
 
from hashlib import md5
 
from hashlib import sha1
 

	
 
from decorator import decorator
 
from tg import config
 

	
 
import kallithea
 
from kallithea.lib.pidlock import DaemonLock, LockHeld
 
from kallithea.lib.utils2 import safe_bytes
 
from kallithea.model import meta
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class FakeTask(object):
 
    """Fake a sync result to make it look like a finished task"""
 

	
 
    def __init__(self, result):
 
        self.result = result
 

	
 
    def failed(self):
 
        return False
 

	
 
    traceback = None # if failed
 

	
 
    task_id = None
 

	
 

	
 
def task(f_org):
 
    """Wrapper of celery.task.task, running async if CELERY_APP
 
    """
 

	
 
    if kallithea.CELERY_APP:
 
        def f_async(*args, **kwargs):
 
            log.info('executing %s task', f_org.__name__)
 
            try:
 
                f_org(*args, **kwargs)
 
            finally:
 
                log.info('executed %s task', f_org.__name__)
 
        f_async.__name__ = f_org.__name__
 
        runner = kallithea.CELERY_APP.task(ignore_result=True)(f_async)
 

	
 
        def f_wrapped(*args, **kwargs):
 
            t = runner.apply_async(args=args, kwargs=kwargs)
 
            log.info('executing task %s in async mode - id %s', f_org, t.task_id)
 
            return t
 
    else:
 
        def f_wrapped(*args, **kwargs):
 
            log.info('executing task %s in sync', f_org.__name__)
 
            try:
 
                result = f_org(*args, **kwargs)
 
            except Exception as e:
 
                log.error('exception executing sync task %s in sync: %r', f_org.__name__, e)
 
                raise # TODO: return this in FakeTask as with async tasks?
 
            return FakeTask(result)
 

	
 
    return f_wrapped
 

	
 

	
 
def __get_lockkey(func, *fargs, **fkwargs):
 
    params = list(fargs)
 
    params.extend(['%s-%s' % ar for ar in fkwargs.items()])
 

	
 
    func_name = str(func.__name__) if hasattr(func, '__name__') else str(func)
 

	
 
    lockkey = 'task_%s.lock' % \
 
        md5(safe_bytes(func_name + '-' + '-'.join(str(x) for x in params))).hexdigest()
 
        sha1(safe_bytes(func_name + '-' + '-'.join(str(x) for x in params))).hexdigest()
 
    return lockkey
 

	
 

	
 
def locked_task(func):
 
    def __wrapper(func, *fargs, **fkwargs):
 
        lockkey = __get_lockkey(func, *fargs, **fkwargs)
 
        lockkey_path = config.get('cache_dir') or config['app_conf']['cache_dir']  # Backward compatibility for TurboGears < 2.4
 

	
 
        log.info('running task with lockkey %s', lockkey)
 
        try:
 
            l = DaemonLock(os.path.join(lockkey_path, lockkey))
 
            ret = func(*fargs, **fkwargs)
 
            l.release()
 
            return ret
 
        except LockHeld:
 
            log.info('LockHeld')
 
            return 'Task with key %s already running' % lockkey
 

	
 
    return decorator(__wrapper, func)
 

	
 

	
 
def get_session():
 
    sa = meta.Session()
 
    return sa
 

	
 

	
 
def dbsession(func):
 
    def __wrapper(func, *fargs, **fkwargs):
 
        try:
 
            ret = func(*fargs, **fkwargs)
 
            return ret
 
        finally:
 
            if kallithea.CELERY_APP and not kallithea.CELERY_EAGER:
 
                meta.Session.remove()
 

	
 
    return decorator(__wrapper, func)
kallithea/lib/markup_renderer.py
Show inline comments
 
@@ -29,103 +29,103 @@ Original author and date, and relevant c
 
import logging
 
import re
 
import traceback
 

	
 
import bleach
 
import markdown as markdown_mod
 

	
 
from kallithea.lib.utils2 import MENTIONS_REGEX, safe_str
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
url_re = re.compile(r'''\bhttps?://(?:[\da-zA-Z0-9@:.-]+)'''
 
                    r'''(?:[/a-zA-Z0-9_=@#~&+%.,:;?!*()-]*[/a-zA-Z0-9_=@#~])?''')
 

	
 

	
 
class MarkupRenderer(object):
 
    RESTRUCTUREDTEXT_DISALLOWED_DIRECTIVES = ['include', 'meta', 'raw']
 

	
 
    MARKDOWN_PAT = re.compile(r'md|mkdn?|mdown|markdown', re.IGNORECASE)
 
    RST_PAT = re.compile(r're?st', re.IGNORECASE)
 
    PLAIN_PAT = re.compile(r'readme', re.IGNORECASE)
 

	
 
    @classmethod
 
    def _detect_renderer(cls, source, filename):
 
        """
 
        runs detection of what renderer should be used for generating html
 
        from a markup language
 

	
 
        filename can be also explicitly a renderer name
 
        """
 

	
 
        if cls.MARKDOWN_PAT.findall(filename):
 
            return cls.markdown
 
        elif cls.RST_PAT.findall(filename):
 
            return cls.rst
 
        elif cls.PLAIN_PAT.findall(filename):
 
            return cls.rst
 
        return cls.plain
 

	
 
    @classmethod
 
    def _flavored_markdown(cls, text):
 
        """
 
        Github style flavored markdown
 

	
 
        :param text:
 
        """
 
        from hashlib import md5
 
        from hashlib import sha1
 

	
 
        # Extract pre blocks.
 
        extractions = {}
 

	
 
        def pre_extraction_callback(matchobj):
 
            digest = md5(matchobj.group(0)).hexdigest()
 
            digest = sha1(matchobj.group(0)).hexdigest()
 
            extractions[digest] = matchobj.group(0)
 
            return "{gfm-extraction-%s}" % digest
 
        pattern = re.compile(r'<pre>.*?</pre>', re.MULTILINE | re.DOTALL)
 
        text = re.sub(pattern, pre_extraction_callback, text)
 

	
 
        # Prevent foo_bar_baz from ending up with an italic word in the middle.
 
        def italic_callback(matchobj):
 
            s = matchobj.group(0)
 
            if list(s).count('_') >= 2:
 
                return s.replace('_', r'\_')
 
            return s
 
        text = re.sub(r'^(?! {4}|\t)\w+_\w+_\w[\w_]*', italic_callback, text)
 

	
 
        # In very clear cases, let newlines become <br /> tags.
 
        def newline_callback(matchobj):
 
            if len(matchobj.group(1)) == 1:
 
                return matchobj.group(0).rstrip() + '  \n'
 
            else:
 
                return matchobj.group(0)
 
        pattern = re.compile(r'^[\w\<][^\n]*(\n+)', re.MULTILINE)
 
        text = re.sub(pattern, newline_callback, text)
 

	
 
        # Insert pre block extractions.
 
        def pre_insert_callback(matchobj):
 
            return '\n\n' + extractions[matchobj.group(1)]
 
        text = re.sub(r'{gfm-extraction-([0-9a-f]{32})\}',
 
                      pre_insert_callback, text)
 

	
 
        return text
 

	
 
    @classmethod
 
    def render(cls, source, filename=None):
 
        """
 
        Renders a given filename using detected renderer
 
        it detects renderers based on file extension or mimetype.
 
        At last it will just do a simple html replacing new lines with <br/>
 

	
 
        >>> MarkupRenderer.render('''<img id="a" style="margin-top:-1000px;color:red" src="http://example.com/test.jpg">''', '.md')
 
        '<p><img id="a" src="http://example.com/test.jpg" style="color: red;"></p>'
 
        >>> MarkupRenderer.render('''<img class="c d" src="file://localhost/test.jpg">''', 'b.mkd')
 
        '<p><img class="c d"></p>'
 
        >>> MarkupRenderer.render('''<a href="foo">foo</a>''', 'c.mkdn')
 
        '<p><a href="foo">foo</a></p>'
 
        >>> MarkupRenderer.render('''<script>alert(1)</script>''', 'd.mdown')
 
        '&lt;script&gt;alert(1)&lt;/script&gt;'
 
        >>> MarkupRenderer.render('''<div onclick="alert(2)">yo</div>''', 'markdown')
 
        '<div>yo</div>'
 
        >>> MarkupRenderer.render('''<a href="javascript:alert(3)">yo</a>''', 'md')
0 comments (0 inline, 0 general)